Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/generic.py

   1 # encoding: utf-8
   2
   3 import os
   4 import re
   5
   6 from .common import InfoExtractor
   7 from ..utils import (
   8     compat_urllib_error,
   9     compat_urllib_parse,
  10     compat_urllib_request,
  11     compat_urlparse,
  12
  13     ExtractorError,
  14 )
  15 from .brightcove import BrightcoveIE
  16
  17
  18 class GenericIE(InfoExtractor):
  19     IE_DESC = u'Generic downloader that works on some sites'
  20     _VALID_URL = r'.*'
  21     IE_NAME = u'generic'
  22     _TESTS = [
  23         {
  24             u'url': u'http://www.hodiho.fr/2013/02/regis-plante-sa-jeep.html',
  25             u'file': u'13601338388002.mp4',
  26             u'md5': u'85b90ccc9d73b4acd9138d3af4c27f89',
  27             u'info_dict': {
  28                 u"uploader": u"www.hodiho.fr",
  29                 u"title": u"R\u00e9gis plante sa Jeep"
  30             }
  31         },
  32     ]
  33
  34     def report_download_webpage(self, video_id):
  35         """Report webpage download."""
  36         if not self._downloader.params.get('test', False):
  37             self._downloader.report_warning(u'Falling back on generic information extractor.')
  38         super(GenericIE, self).report_download_webpage(video_id)
  39
  40     def report_following_redirect(self, new_url):
  41         """Report information extraction."""
  42         self._downloader.to_screen(u'[redirect] Following redirect to %s' % new_url)
  43
  44     def _test_redirect(self, url):
  45         """Check if it is a redirect, like url shorteners, in case return the new url."""
  46         class HeadRequest(compat_urllib_request.Request):
  47             def get_method(self):
  48                 return "HEAD"
  49
  50         class HEADRedirectHandler(compat_urllib_request.HTTPRedirectHandler):
  51             """
  52             Subclass the HTTPRedirectHandler to make it use our
  53             HeadRequest also on the redirected URL
  54             """
  55             def redirect_request(self, req, fp, code, msg, headers, newurl):
  56                 if code in (301, 302, 303, 307):
  57                     newurl = newurl.replace(' ', '%20')
  58                     newheaders = dict((k,v) for k,v in req.headers.items()
  59                                       if k.lower() not in ("content-length", "content-type"))
  60                     return HeadRequest(newurl,
  61                                        headers=newheaders,
  62                                        origin_req_host=req.get_origin_req_host(),
  63                                        unverifiable=True)
  64                 else:
  65                     raise compat_urllib_error.HTTPError(req.get_full_url(), code, msg, headers, fp)
  66
  67         class HTTPMethodFallback(compat_urllib_request.BaseHandler):
  68             """
  69             Fallback to GET if HEAD is not allowed (405 HTTP error)
  70             """
  71             def http_error_405(self, req, fp, code, msg, headers):
  72                 fp.read()
  73                 fp.close()
  74
  75                 newheaders = dict((k,v) for k,v in req.headers.items()
  76                                   if k.lower() not in ("content-length", "content-type"))
  77                 return self.parent.open(compat_urllib_request.Request(req.get_full_url(),
  78                                                  headers=newheaders,
  79                                                  origin_req_host=req.get_origin_req_host(),
  80                                                  unverifiable=True))
  81
  82         # Build our opener
  83         opener = compat_urllib_request.OpenerDirector()
  84         for handler in [compat_urllib_request.HTTPHandler, compat_urllib_request.HTTPDefaultErrorHandler,
  85                         HTTPMethodFallback, HEADRedirectHandler,
  86                         compat_urllib_request.HTTPErrorProcessor, compat_urllib_request.HTTPSHandler]:
  87             opener.add_handler(handler())
  88
  89         response = opener.open(HeadRequest(url))
  90         if response is None:
  91             raise ExtractorError(u'Invalid URL protocol')
  92         new_url = response.geturl()
  93
  94         if url == new_url:
  95             return False
  96
  97         self.report_following_redirect(new_url)
  98         return new_url
  99
 100     def _real_extract(self, url):
 101         parsed_url = compat_urlparse.urlparse(url)
 102         if not parsed_url.scheme:
 103             self._downloader.report_warning('The url doesn\'t specify the protocol, trying with http')
 104             return self.url_result('http://' + url)
 105
 106         try:
 107             new_url = self._test_redirect(url)
 108             if new_url:
 109                 return [self.url_result(new_url)]
 110         except compat_urllib_error.HTTPError:
 111             # This may be a stupid server that doesn't like HEAD, our UA, or so
 112             pass
 113
 114         video_id = url.split('/')[-1]
 115         try:
 116             webpage = self._download_webpage(url, video_id)
 117         except ValueError:
 118             # since this is the last-resort InfoExtractor, if
 119             # this error is thrown, it'll be thrown here
 120             raise ExtractorError(u'Invalid URL: %s' % url)
 121
 122         self.report_extraction(video_id)
 123         # Look for BrightCove:
 124         m_brightcove = re.search(r'<object.+?class=([\'"]).*?BrightcoveExperience.*?\1.+?</object>', webpage, re.DOTALL)
 125         if m_brightcove is not None:
 126             self.to_screen(u'Brightcove video detected.')
 127             bc_url = BrightcoveIE._build_brighcove_url(m_brightcove.group())
 128             return self.url_result(bc_url, 'Brightcove')
 129
 130         # Start with something easy: JW Player in SWFObject
 131         mobj = re.search(r'flashvars: [\'"](?:.*&)?file=(http[^\'"&]*)', webpage)
 132         if mobj is None:
 133             # Broaden the search a little bit
 134             mobj = re.search(r'[^A-Za-z0-9]?(?:file|source)=(http[^\'"&]*)', webpage)
 135         if mobj is None:
 136             # Broaden the search a little bit: JWPlayer JS loader
 137             mobj = re.search(r'[^A-Za-z0-9]?file["\']?:\s*["\'](http[^\'"&]*)', webpage)
 138         if mobj is None:
 139             # Try to find twitter cards info
 140             mobj = re.search(r'<meta (?:property|name)="twitter:player:stream" (?:content|value)="(.+?)"', webpage)
 141         if mobj is None:
 142             # We look for Open Graph info:
 143             # We have to match any number spaces between elements, some sites try to align them (eg.: statigr.am)
 144             m_video_type = re.search(r'<meta.*?property="og:video:type".*?content="video/(.*?)"', webpage)
 145             # We only look in og:video if the MIME type is a video, don't try if it's a Flash player:
 146             if m_video_type is not None:
 147                 mobj = re.search(r'<meta.*?property="og:video".*?content="(.*?)"', webpage)
 148         if mobj is None:
 149             # HTML5 video
 150             mobj = re.search(r'<video[^<]*(?:>.*?<source.*?)? src="([^"]+)"', webpage, flags=re.DOTALL)
 151         if mobj is None:
 152             raise ExtractorError(u'Invalid URL: %s' % url)
 153
 154         # It's possible that one of the regexes
 155         # matched, but returned an empty group:
 156         if mobj.group(1) is None:
 157             raise ExtractorError(u'Invalid URL: %s' % url)
 158
 159         video_url = mobj.group(1)
 160         video_url = compat_urlparse.urljoin(url, video_url)
 161         video_id = compat_urllib_parse.unquote(os.path.basename(video_url))
 162
 163         # here's a fun little line of code for you:
 164         video_extension = os.path.splitext(video_id)[1][1:]
 165         video_id = os.path.splitext(video_id)[0]
 166
 167         # it's tempting to parse this further, but you would
 168         # have to take into account all the variations like
 169         #   Video Title - Site Name
 170         #   Site Name | Video Title
 171         #   Video Title - Tagline | Site Name
 172         # and so on and so forth; it's just not practical
 173         video_title = self._html_search_regex(r'<title>(.*)</title>',
 174             webpage, u'video title', default=u'video', flags=re.DOTALL)
 175
 176         # video uploader is domain name
 177         video_uploader = self._search_regex(r'(?:https?://)?([^/]*)/.*',
 178             url, u'video uploader')
 179
 180         return [{
 181             'id':       video_id,
 182             'url':      video_url,
 183             'uploader': video_uploader,
 184             'upload_date':  None,
 185             'title':    video_title,
 186             'ext':      video_extension,
 187         }]